The goals of the proposed study are to explore developmental changes in language acquisition and theory of mind in well-defined group of autistic and matched control subjects. 60 autistic children ranging in IQ from about 40 to 100, and in age from 4 to 14 at the start of the study (divided into three cohorts)will be followed longitudinally in a cohort sequential design over a span of four to five years. The autistic subjects will be matched at the start of the study to a group of 60 mentally retarded or language/learning disabled subjects on age, gender, SES, IQ and language level, as assessed by a receptive vocabulary measure. Yearly assessments of language, drawing on evaluations that will be conducted by the Clinical and Administrative Core of the program project, will provide indices of overall language development, and component measures of morphosyntactic, lexical- semantic and pragramtic/discourse functioning. For this project, the subjects will also be seen annually, at which time they will be administered an experimental battery of theory of mind tasks and a set of language tasks designed to est specific knowledge of linguistic complementation. The data derived from these tasks (and other Core data available on the subjects) will be used to explore both general and specific developmental relationship between these two broad domains: language and theory of mind. The particular aims that guide this study include: (1) To investigate developmental changes in language in the two main groups of subject. We hypothesize that the autistic group will show less significant overall development change, exhibit a more uneven patterns of relationships among the components of language (morphosyntax, lexicon, pragmatics), and that IQ will be an important predictor of rate of developmental change. (2) To investigate developmental changes in theory of mind in the two main groups of subjects. Again, we hypothesize that the autistic subjects will show less significant overall developmental change, but that only for the mentally retarded subjects will IQ be a significant predictor of performance. (3) to investigate the developmental relationships between both general, and specific aspects of theory of mind and language (pragmatics; linguistic complementation) in the two groups of subjects. We hypothesize that these two domains are closely related over the course of development for all the subjects in the study.